OBJECTIVE

To investigate the use of prenatal maternal serum screening results for Down syndrome, for the prediction of low (<2500 g) and very low (<1500 g) birthweight.

METHODS

Record linkage of maternal serum screening results with the corresponding birth records.

METHODS

42 259 women whose pregnancies had been screened for the risk of Down syndrome.

METHODS

Three East London maternity units, between February 1989 and August 1998.

RESULTS

Estimates were made of the effectiveness of single markers only for the prediction of low birthweight, and of multiple markers together with mother's weight and smoking habit. As reported previously, high levels of the single markers alpha-fetoprotein and total human chorionic gonadotrophin, inhibin A, and low levels of unconjugated oestriol were associated with low birthweight. However, the best prediction was obtained when multiple serum markers comprising alpha-fetoprotein, unconjugated oestriol, and inhibin A were used in combination together with mother's weight and adjustment for smoking habit. For a false-positive rate of 5%, this combination predicted 23% of low birthweight and 39% of very low birthweight babies, possibly the best method of prediction to date.

CONCLUSIONS

Prediction of low birthweight derived from Down syndrome screening could be used, for little extra cost, to advise on place of delivery or to select candidates for randomised clinical trials of low birthweight prevention.

The efficacy of five different methods for detection of fetoplacental distress has been compared in 265 risk pregnancies. The tests were urinary oestriol, urinary pregnanediol, human placental lactogen and heat-stable alkaline phosphotase in the serum and phonocardiotokographie. 13 fetuses died antepartum, 47 of surviving infants were growth-retarded. The gestational age was calculated from clinical dates, serial ultrasonic measurements and immediate postnatal studies. In the I part the methods and the analysis of the results are described.

OBJECTIVE

To evaluate the usefulness of the two maternal serum markers, human chorionic gonadotrophin (hCG) and unconjugated oestriol (uE3), in the prenatal diagnosis of trisomy 18.

METHODS

Retrospective evaluation of uE3 and hCG levels at mid-trimester in cases ot trisomy 18 pregnancies identified from a series of women screened for Down's syndrome.

METHODS

From a series of 53,893 women screened in the antenatal centre of University Hospital of Caen (France), 22 cases of trisomy 18 were diagnosed either after amniocentesis for maternal age, elevated risk of Down's syndrome, or fetal abnormalities and/or growth retardation on ultrasound assessment, or after birth. In addition, ll cases of trisomy 18 identified prenatally in two other centres were included.

RESULTS

Individual hCG and uE3 levels for pregnancies with trisomy 18 were significantly lower than in unaffected pregnancies: mean hCG was 0.62 multiples of the median (MoM) and median hCG was 0.5 MoM. uE3 was a much more effective marker than hCG. Mean uE3 was 0.40 MoM and median uE3 was 0.37 MoM. It was observed that screening for trisomy 18 based on selection for amniocentesis with cut-off values of 0.55 for hCG and 0.60 for uE3 would lead to a detection rate of 48% for 0.8% false positive rate. Using cut-off values of 0.70 MoM for each one of the two markers would detect 79% of cases of trisomy 18 with 3% false positive rate.

CONCLUSIONS

Our results confirm that low hCG and uE3 levels observed in the mid-trimester are predictive of an increased risk for trisomy 18. Since most fetuses with trisomy 18 exhibit morphological abnormalities which should be detected following a careful ultrasonographic examination, biochemical screening could help in the detection of those anatomical defects in selecting for scanning a group of high risk women.

A method for the determination of oestriol in pregnancy urine by normal-phase high-performance liquid chromatography with electrochemical detection is described. A large-volume wall-jet cell with an Ag-Ag+ reference electrode was used as the detector system. The limit of detection obtained is comparable to that of electrochemical detection following reversed-phase liquid chromatography. One of the advantages of electrochemical detection with normal-phase systems is that adsorption problems are minimized.

Maternal sera (MS) taken from 1396 women prior to chorionic villus sampling at 9-12 menstrual weeks were assayed for unconjugated oestriol (uE3) and alpha-fetoprotein (AFP). Median levels increased by 41 and 26 per cent per week respectively in normal pregnancies. There were 32 pregnancies with a chromosome abnormality. The median MS uE3 and AFP were 0.73 and 0.75 multiples of the median (MoM) respectively in the ten cases of Down's syndrome (DS) but not decreased in the other abnormalities. These results suggest that both uE3 and AFP may be useful in identifying DS in the first trimester. Additional prospective studies are needed to confirm these findings.

The serum markers human chorionic gonadotrophin (hCG), alpha-fetoprotein (AFP) and unconjugated oestriol (uE3) in 606 rhesus-negative (Rh-) women have been compared with 18 960 controls. There were no significant differences in the distribution of hCG and AFP between these two cohorts. However the uE3 values in Rh- women were significantly lowered (median: 0.85 MoM).

Plasma levels of unconjugated oestriol and urinary excretion of conjugated oestriol were measured in ten healthy women with uncomplicated late pregnancy and with normal renal function. Considerably larger differences within or between days were noted for the plasma levels than for the urinary excretion of oestriol. Significant correlation between plasma levels and urinary excretion was found in only three out of ten patients. Due to the large variations within and between days, the estimation of unconjugated oestriol in plasma might be unsuitable as a substitute for the estimation of urinary conjugated oestriol in the supervision of complicated pregnancies.

A cohort of 1238 consecutive women seen early in pregnancy for accurate assessment of gestation had urinary excretion of oestriol measured at 30, 35 and 40 weeks; 92 had a low urinary oestriol excretion confirmed in hospital on at least one occasion. The women in the low-oestriol group were matched for racial origin and maternal age with 90 women in the study population who had persistently normal urinary oestriol excretion values (control group) and their infants were compared. Three infants born in the low-oestriol group and two in the control group subsequently died; 85 of the 89 (95.5%) survivors in the low-oestriol group and 84 of 88 (95.5%) control children were assessed by a paediatrician and a psychologist at the age of 6 years. Anthropometric measurements of children in the low-oestriol group at 6 years of age did not differ significantly from controls and their performance on most psychological tests was not significantly different. The subgroup of 25 children born to women with persistently low oestriol excretion during pregnancy showed a significant reduction in weight and head circumference at 6 years compared with the values in the control group and in the other children in the low-oestriol group. The children born to women with low oestriol excretion during pregnancy showed no increased evidence of developmental, neurological or physical defects at 6 years of age.

Urine oestriol values from 327 pregnancies with antenatal complications were measured between 30 weeks gestation and term. Mean values were much lower than those obtained from pregnancies without complications. When infants were light for dates or developed evidence of neonatal morbidity, oestriol values were more likely to have been below the normal range than in pregnancies resulting in deliveries of normal infants. There was no evidence of severe fetal jeopardy in most pregnancies with oestriol values below normal. Perinatal death was consistently predicted when urine oestriol values lay below a line joining 35 mumol/day at 30 weeks and 65 mumol/day at term.

A specific radioimmunoassay has been developed for measuring oestriol-16alpha glucuronide in pregnancy plasma using a highly specific antiserum. The specificity of this antiserum has been assessed by the 50% displacement method and by measuring oestriol-16alpha-glucuronide concentrations in pregnancy plasma samples, both with and without chromatography. The antiserum was found to have Ka of 1.3 X 10(10) M-1. The assay had a sensitivity of 5 pg (least quantity of oestriol-16alpha-glucuronide distinguishable from the zero point, p less than 0.05), an intra-assay variation of 6.07% and an inter-assay variation of 3.36%. The assay was used to measure plasma oestriol-16alpha-glucuronide concentrations throughout eight pregnancies, five of which were normal and three had some defined abnormality.

Saliva specimens were collected at weekly intervals during the third trimester of 61 uncomplicated pregnancies. Oestriol concentrations, measured in samples collected in the early morning from fasting, recumbent subjects, were used to construct a provisional normal range. This proved to be higher than the corresponding normal range reported from another laboratory applying the same method to samples obtained at random from a different population of ambulant subjects. Interlaboratory variation in the assay of salivary oestriol was not sufficient to account for the difference in normal range. It is concluded that consideration should be given to the timing of sample collection if salivary oestriol concentration is to be monitored as a possible means of assessing fetoplacental function. A normal range for salivary oestriol concentration versus gestational age obtained from women sampled randomly may not be appropriate for monitoring patients with 'at-risk' pregnancies who are normally bed-rested.

Six postmenopausal women were treated at two ocasions with an oral dose of, respectively, 2 mg and 6 mg oestriol per day for 14 days; at least one month interval separated the two treatment periods. Blood was collected for prolactin measurement (homologous human radioimmunoassay) every other day during each treatment period as well as during a 14 days control period in the same subjects. Mean prolactin level during oestriol administration, regardless of the dose, was statistically not significantly different from that observed during the control period.